Pancreatic Cancer Vaccine Falls Short in Phase III Trial

George Fisher, MD, PhD

The investigational pancreatic cancer vaccine algenpantucel-L failed to improve overall survival (OS) versus standard of care in the phase III IMPRESS trial, according to a statement from NewLink Genetics, the developer of the immunotherapy.

The median OS was 27.3 months when combining algenpantucel-L with standard of care (chemotherapy with or without radiation) versus 30.4 months with standard of care alone. For both groups combined, the OS from the time of randomization was 29.3 months.

"Immunotherapy is rapidly establishing a role in the management of multiple malignancies," George Fisher, MD, PhD, professor of Medicine at Stanford University and an investigator on the IMPRESS study, said in a statement. "The median overall survival of 29.3 months in this study represents a significant increase compared with prior trials and may be due to multiple factors, including the emergence of more effective treatment regimens for recurrent or metastatic disease. Although a negative study, these results represent an important and meaningful contribution to the understanding of the modern treatment of resected pancreatic cancer."

The open-label IMPRESS trial included 722 patients with pancreatic cancer who had undergone resection. Patients were randomized to receive an adjuvant regimen of either algenpantucel-L (up to 18 immunizations of 300 million cells) plus gemcitabine (1000 mg/ m²/day, once a week for 3 weeks) with or without 5-FU (200-250 mg/m²/day over 5.5 weeks) with radiation or gemcitabine with or without 5-FU chemoradiation.

The median age of patients was 65 years, 52% of patients were male, and 80% of tumors were resected from the head of the pancreas. OS was the primary endpoint, with secondary outcome measures including disease-free survival (DFS) and determining the mechanism of any observed antitumor effect.

The difference in long-term survival between the 2 arms was not statistically significant. The 3-year survival rate was 42.1% for the immunotherapy arm and 41.4% with standard of care alone. The 4-year survival rates were 32.7% and 32.6%, respectively.

"We are deeply disappointed for patients that the IMPRESS phase III study was not successful," Nicholas N. Vahanian, MD, president and chief medical officer of NewLink Genetics, said in a statement. “We want to extend our sincere appreciation to all the patients, caregivers, investigators, research nurses, employees, and others who contributed to the study.”

“In light of these negative results, our scientific and clinical teams will focus on other promising opportunities in our pipeline," Charles Link, Jr, MD, chairman and CEO of NewLink Genetics, added in a statement. "Our lead projects focus on our IDO checkpoint inhibitor technology employing indoximod and GDC-0919.”

The IMPRESS trial was launched following positive findings in a phase II study in which patients with resected pancreatic cancer received algenpantucel-L in addition to chemotherapy and chemoradiation in the adjuvant setting.1 In this study, 73 patients were enrolled, with 69 meeting the criteria.

At a median follow-up of 21 months, the 1-year DFS was 62%, and the 12-month OS was 86%. The median DFS was 14.1 months.

The most common grade 1 and 2 adverse events (AEs) were induration, fatigue, and injection site reaction. Twelve percent of patients experienced grade 3 AEs and no grade 4 events; grade 3 events included fatigue, injection site reaction, pain, and lymphopenia.

Algenpantucel-L is also being examined in the open-label phase III randomized PILLAR trial (NCT01836432), which completed its enrollment of over 300 patients in December 2015. In the study, patients with borderline resectable or locally advanced unresectable pancreatic cancer were randomized to FOLFIRINOX or gemcitabine/nab-paclitaxel with or without algenpantucel-L.

1. Hardacre JM, Mulcahy M, Small W, et al. Addition of algenpantucel-L immunotherapy to standard adjuvant therapy for pancreatic cancer: a phase 2 study. J Gastrointest Surg. 2013;17(1):94-100.